Report

# Conversion of Zinc Oxide Nanobelts into Superlattice-Structured Nanohelices

See allHide authors and affiliations

Science  09 Sep 2005:
Vol. 309, Issue 5741, pp. 1700-1704
DOI: 10.1126/science.1116495

## Abstract

A previously unknown rigid helical structure of zinc oxide consisting of a superlattice-structured nanobelt was formed spontaneously in a vapor-solid growth process. Starting from a single-crystal stiff nanoribbon dominated by the c-plane polar surfaces, an abrupt structural transformation into the superlattice-structured nanobelt led to the formation of a uniform nanohelix due to a rigid lattice rotation or twisting. The nanohelix was made of two types of alternating and periodically distributed long crystal stripes, which were oriented with their c axes perpendicular to each other. The nanohelix terminated by transforming into a single-crystal nanobelt dominated by nonpolar ($Math$) surfaces. The nanohelix could be manipulated, and its elastic properties were measured, which suggests possible uses in electromechanically coupled sensors, transducers, and resonators.

Helical structures have been observed for a number of inorganic materials. For example, carbon nanotube coils (1) are created when paired pentagon-heptagon atomic rings arrange themselves periodically within the hexagonal carbon network (2). Formation of nanosprings of amorphous silica has been thought to occur through a perturbation during the growth of a straight nanowire (3). Helical structures of SiC are proposed to be a screw-dislocation-driven growth process (4). By using the difference in surface stress on the two surfaces, rings and tubes of strained bilayer thin films, such as Si/SiGe, have been made (5).

Zinc oxide, an important semiconductive and piezoelectric material with several applications in optoelectronics and sensing, can form single-crystal nanosprings (6), nanorings (7), and nanobows (8) by bending or folding polar-surface-dominated nanobelts. These growth processes are dominated by a minimization of the electrostatic energy contributed by the cation- and anion-terminated surfaces of this ionic material, and their shapes are determined by balancing the electrostatic energy with the elastic deformation energy. The structural configurations of ZnO nanoobjects (9) could form the basis for fabricating nanoscale sensors, transducers, and resonators.

We now report a distinctive helical structure of ZnO made of a superlattice-structured nanobelt that is formed spontaneously in a solid-vapor growth process. The superlattice nanobelt is a periodic, coherent, epitaxial, and parallel growth of two alternating nanostripes of ZnO crystals oriented with their c axes perpendicular to each other. The role played by polar surfaces in the structural transformation is demonstrated, and the nanohelix has been manipulated by atomic force microscopy (AFM) to measure its elastic properties.

The ZnO nanohelices were grown with high reproducibility by a vapor-solid process (10) and by using temperature to control growth kinetics. The experimental setup consists of a horizontal high-temperature tube furnace, an alumina tube, a rotary pump system, and a gas-controlling system. First, 2 g of commercial ZnO powder (Alfa Aesar, Ward Hill, MA) were compacted, loaded into an alumina boat, and positioned at the center of the alumina tube as the source material. The system was prepumped to ∼2 × 10–2 mbar, and the ramp rate was controlled at 20°C/min to 25°C/min when the temperature was raised from room temperature to 800°C. The furnace was then held at 800°C for 20 min, and the temperature was ramped at 20°C/min from 800° to 1400°C. When the temperature reached 1000°C, argon was introduced as a carrier gas to raise the pressure from ∼2 × 10–2 mbar to the desired synthesis pressure of 200 to 250 mbar within ∼2.5 min. The solid-vapor deposition was carried out at 1400°C for ∼2 hours under a pressure of 200 to 250 mbar. The argon carrier gas was kept at a flow rate of 50 sccm (standard cubic centimeters per minute). The as-grown nanohelices of ZnO were deposited onto a polycrystalline Al2O3 substrate at a local temperature of 700° to 800°C and characterized by a variety of microscopes, including scanning electron microscopy (SEM) [LEO 1530 FEG (Oberkochen, Germany) at 5 and 10 kV], high-resolution transmission electron microscopy (TEM) [Hitachi HF-2000 at 200 kV and JEOL 4000EX (Tokyo, Japan) at 400 kV], AFM [Molecular Force Probe MFP-3D from Asylum Research (Santa Barbara, CA)], and focused ion beam microscopy (FIB) (FEI NOVA 200).

The as-synthesized sample was composed of up to 10% of freestanding nanohelices (fig. S1). High-resolution SEM images show the high degree of uniformity of the nanohelices, which can be left- (Fig. 1, A and C) or right-handed (Fig. 1, B and D). Their population ratio is nearly 1:1 (fig. S2). As shown in Fig. 1, the nanohelix diameters, widths, and pitch distances range between 300 to 700 nm, 100 to 500 nm, and 500 to 2500 nm, respectively (11). The length of the nanohelix can be as long as 100 μm. Energy dispersive x-ray spectroscopy analysis indicates that the nanorings are ZnO without detectable impurity.

Low-magnification TEM images of the nanohelices (Fig. 2, A and B) revealed the crystalline structure and winding shape of the nanobelts. The electron diffraction (ED) pattern, acquired from a single period of a nanohelix, shows mirror symmetry. This mirror symmetry is caused by superposition of several ZnO diffraction patterns rotated around the axial direction of the nanohelix axial direction of the nanohelix. A dark-field TEM image recorded from a segment of a nanobelt (Fig. 2C) shows that the nanobelt has a periodic superlattice structure across its entire width, with a uniform periodicity of ∼3.5 nm. With a 5° offset, the stripes are nearly parallel to the nanobelt growth direction and run along the length of the nanohelix (fig. S3).

From crystallography, the wurtzite-structured ZnO crystal is described schematically as a number of alternating planes of tetrahedrally coordinated O2– and Zn2+ ions stacked alternatively along the c axis. The oppositely charged ions produce positively charged (0001)-Zn2+ and negatively charged $Math$ polar surfaces (c planes) (7, 9). High-resolution TEM revealed the superlattice structure of the nanobelt with a periodic alternating arrangement of two types of stripes (Fig. 2D), which are oriented epitaxially in the following orientations: Stripe I has a growth direction of $Math$, with top and bottom surfaces ±(0001) (polar surfaces); stripe II has a growth direction of $Math$, with top and bottom surfaces $Math$ (nonpolar surfaces). The corresponding ED pattern of the nanobelt shows that the growth directions of the two types of stripes have a small rotation of α ≈ 4° (Fig. 2E), indicating that the adjacent stripes rotate in the plane of the nanobelt for the same magnitude of the angle when forming the superlattice. The angles between the ED spots also slightly deviate from the expected result of a large crystal and suggest rigid lattice twist in some atomic planes. The two stripes are coherent by aligning $Math$ of stripe I with (0002) of stripe II at a lattice mismatch of 2.9% and ($Math$) of stripe I with ($Math$) of stripe II without lattice mismatch, so the interface energy is likely to be low. Viewing from the projected direction of the incident electron beam, we aligned the two stripes by facing ($Math$) of stripe I parallel to the ($Math$) of stripe II and aligned the [0001] of stripe I with [$Math$] of stripe II. The true interface is inclined with respect to the direction of the electron beam, so that the contrast at the interface in the high-resolution TEM image is not very sharp (Fig. 2F). On the basis of the orientation relation presented above, the experimental image can be simulated by using dynamic ED theory. The result is inserted in Fig. 2F (11), showing quantitative agreement with the observation. Thus, the interface for stripe I is likely to be ($Math$) (12).

The initiation of the nanohelix is a direct result of superlattice formation. From an SEM image of a nanohelix with both ends (Fig. 3A), a straight stiff nanoribbon is directly connected to the starting point of a nanohelix. TEM examination shows that the joining point between the nanoribbon and the nanohelix is the place where the superlattice is created (Fig. 3B). The stiff nanoribbon is a single crystal object dominated by large c-plane surfaces (Fig. 3C). The superlattice is formed spontaneously and abruptly across the entire width of the nanoribbon. A striking feature is that the width of the nanobelt increases and its thickness decreases significantly after transforming into the superlattice structure (Fig. 3B and fig. S4). The stripes hold the same orientation relation as defined in Fig. 2E [see the selected-area ED (SAED) pattern inserted in Fig. 3C]. An enlarged local area shows the coexistence and spontaneous formation of two periodically arranged, alternating stripes at the transverse interface where the structural transformation occurred abruptly (Fig. 3D). At the transverse interface, the lattice mismatch of 2.9% between ($Math$) of the nanoribbon with (0002) of stripe II should create interface mismatch edge dislocations with a Burgers vector of [$Math$], which are expected to be separated by 3.45 nm. In fact, the periodicity of the superlattice is ∼3.5 nm, and mismatch edge dislocations have been identified at the turning point between the nanoribbon and the superlattice (circled areas in Fig. 3D). The site of the mismatch dislocation could be the nucleation site of stripe II. The superior uniformity in periodicity of the superlattice is most likely to be a result of the lattice mismatch as defined by crystal structure, and rigid crystal twisting gives rise to the strain contrast observed in Fig. 3B and even in the high-resolution TEM image in Fig. 3D.

The termination of the nanohelix growth is caused by a complete conversion of the remaining polar surfaces into nonpolar surfaces and an increase in nanoribbon thickness. From the TEM image and the corresponding SAED patterns recorded from a single nanohelix (Fig. 3E), the formation of the nanohelix is a result of transforming a fully c-plane-dominated stiff nanoribbon (starting point) into a superlattice-structured, partially polar-surface-dominated flexible nanobelt (middle section). The starting stiff nanobelt is thick (∼40 nm) and straight (fig. S4). After transforming into the superlattice-structured nanobelt, the thickness of the nanobelt is greatly reduced to less than 20 nm, and its width is substantially increased (fig. S4). Lastly, the ending of the nanohelix is a result of abruptly transforming the partially c-plane-dominated, superlattice-structured nanobelt into a single-crystal nanoribbon dominated by nonpolar surfaces at the finishing end (Fig. 3E), where the corresponding dark-field TEM image and the high-resolution TEM image (Fig. 3F) indicate that the top and bottom surfaces of the nanoribbon are the nonpolar ±($Math$) surfaces. The thickness of the nanoribbon increases significantly as well after the transformation.

The formation of the nanohelix is a rigid structural alteration caused by the alternating stripes in the superlattice, and the formation of a superlattice is likely to be a polar charge–induced structural transformation (see the model in Fig. 3G). In the synthesis, a prepumping to 2 × 10–2 mbar creates the condition for growing polar surface–dominated nanoribbons (13). The polar charges are preserved on the surface during growth under conditions of high temperature and inert environment (14). As reported previously (6, 7), the existence of the polar surfaces results in the formation of ZnO nanosprings and nanorings. If the width of the nanoribbon is rather large (Fig. 3B), the electrostatic energy of the polar surfaces diverges with the extension of its length. A possible way of reducing the electrostatic energy is to transform the polar surface into a nonpolar surface, provided the interface mismatch energy is reasonably low, forming superlattice-structured stripes that are nearly parallel to the growth direction of the nanobelt. The introduction of argon gas at a later stage during the growth may be the initiation of superlattice formation. The switching in the c axis may indicate the formation of piezoelectric domains. The width of the nanobelt increases after the structural transformation from a single-crystal, c-plane-dominated stiff nanoribbon (in green) into a superlattice-structured, flexible nanobelt; the stripes are not exactly parallel to the growth direction of the nanobelt, and a small in-plane rotation for an angle of α exists between the growth directions of the two adjacent stripes, which induces the geometrical curvature required for rigidly twisting its shape. The polar charges remaining on the surfaces of stripe I also help to bend the nanobelt into a ring structure, especially with the decrease of the nanobelt thickness (15). Thus, the accumulation of rigid structural rotation or twist across the width of the superlattice nanobelt and a continuous growth of the stripes along the length of the nanobelt result in the formation of a helical structure, which could be left-handed or right-handed. The winding of the nanobelt is terminated after the nanobelt fully transforms from a partial-polar-surface-dominated and superlattice-structured nanoobject into a single-crystal nanoribbon dominated by nonpolar surface (in blue), which is likely to be caused by a change in growth kinetics. The single-crystal nanoribbon is a straight structure without rigid lattice alteration or surface polar charges; thus, it will not continue to wind but will extend as a straight ribbon.

Manipulation of a nanohelix with AFM and FIB shows that it has a rigid structure. A nanohelix has been sectioned by FIB into a short segment of one period (Fig. 4A), but the shape of the segment is still preserved and shows no visible relaxation. A nanohelix has also been cut by an AFM tip at two consecutive turns (Fig. 4C), and the preservation of the rigid shape by a single-turn helix after cutting suggests little, if any, stored elastic energy (16).

The transverse spring constant of the nanohelix, a quantity for describing the stiffness of the nanohelix perpendicular to its axial direction, has been measured by AFM with use of the force-displacement (F-Z) curve (17). Through the static compression of a nanohelix lying on a silicon substrate with an AFM tip (Fig. 4B) (fig. S5) and careful calibration of the sensitivity, S, of the photodetector (11, 18), which is the inverse of the slope of the F-Z curve (fig. S6), the transverse spring constant of the nanohelix, KHelix, was found to be (11) 3.9, 4.6, 4.5, and 5.3 N/m for one to four turns, respectively, of the nanohelix as labeled in Fig. 4B. The measured KHelix is near the theoretically estimated transverse spring constant of 4.2 N/m (11), and so the elastic modulus, E, values of the nanohelix derived from the measured spring constant (11) are 42, 49, 48, and 57 GPa for the four turns, respectively, which agrees well with the elastic modulus measured for straight ZnO nanobelts by mechanical resonance with the use of in situ TEM (19).

The perfect helix we observed is of great interest not only for understanding the growth behavior of polar-surface–driven growth processes in the wurtzite system, such as GaN, AlN, and InN, but also for investigating fundamental physics and optical phenomena. The piezoelectric and semiconducting properties of ZnO suggest that the nanohelix could be a fundamental unit for investigating electromechanically coupled nanodevices by using the superlattice piezoelectric domains. The nanohelix is likely to have important applications in sensors, transducers, resonators, and photonics.

Supporting Online Material

Materials and Methods

Figs. S1 to S6

View Abstract